Theory: perhaps the letters do not have phonemic content?

[ololololo]

The problem with the theory of artificial language is that there is no proper methodology. If this is indeed an old conlang, then logically, the content of VMS is probably mystical, but more so with an alchemical twist. Therefore, it is more likely to be an alchemical language rather than a philosophical one.
If this is the case, then I assume that this language is very limited and incomplete compared to other languages, but it is also rigidly functional, meaning that it can be used to describe what is described in VMS.

Be careful, this is a sure way towards a Procedural Structural-Semantical Framework for Functional Transformative Alchemical Computation with Robust Verifiable Results achieving Positive Coverage of Over 76% of the Text (link to Zenodo).

Fortunately, that's not a problem for me.  
Moreover, I don't think it's possible to translate this language; we can only infer its properties.